How to set and decide appropriate df0 and S0 for distribution of genetic variance prior? #82
Description
Hi,
I'm using the BGLR package for genomic prediction (Bayes A, BayesB) and want to better understand how to set hyperparameters for the distribution of genetic variance, particularly using the df0 (degrees of freedom) and S0 (scale) parameters. I notice that changes to df0 and S0 seem to influence the residual variance, but not the genetic component. How should I decide appropriate values for df0 and S0, and how to set them for genetic variance?
Any clarification on how to choose and apply these hyperparameters for the distribution of genetic variance appropriately would be greatly appreciated.